CompuServe Thread

#Help!

3 messages in this thread
#12506From: Gene NeronJul 25, 1991 11:01 PM
I'm having some strange problems that maybe you can help me with. I have a 2500/20 with a 2090A controller, XT bridgeboard and 5 megs of memory (4 of them 32 bit memory). I just installed a 105 meg SCSI drive in my system, and now some funny things are happening. I thought it might be related to the new hard disk,but I took the disk out and I have the same problems. Here are my symptoms: 1) Setcpu crashes the system. When I type setcpu fastrom, it pauses for a few seconds, then a "whooom" sound comes out of the speakers, the screen flickers and goes black, and the system is dead. When I reboot, the clock has lost the time completely. The setclock load returns <unset> <unset> <unset> and I have to manually set it. Setcpu used to work fine. 2) When I make of copy of the Empty drawer using workbench, the system gurus with a 81000005, which is a corrupt memory list. I can duplicate the drawer with the CLI, but Workbench always crashes. I tried using nofastmem to disable all fast memory, and then I can use WB to duplicate the empty drawer. I tried putting nofastmem at the beginning of my startup sequence and then again at the end to force everything into chip ram. When I do this and use WB to duplicate the Empty drawer, it works, except somehow all my fast memory gets allocated to something afterwards, ie, I have no fast memory at all after the operation is complete. I didn't have any of these problems with my old disk, and putting everything back the way it was (ie, completely removing the new disk and replacing the old) does not get rid of the problems, so I apparantly have disturbed something in the system but I don't know what. I thought I might have a bad memory chip so I tried running some memory diagnostics I downloaded from here. One reports no errors, the other gurus with a "B". One problem I have in testing memory is that the system allocates about 87K of fast memory even if I put nofastmem as the first command in the startup-sequence file, so that portion of fast ram doesn't get tested. Any way around this? Do you have any ideas about what might be wrong? Please help! Thanks! Gene
#12526From: Michael GerardJul 26, 1991 6:26 PM
Gene: I can't help too much, but I just got a call from a friend who had a similar problem with a GVP 68030 (A3001). It worked fine with the original 4 meg of 32 bit ram, but a short time after he added an additional 4 meg, the same type symptons you describe showed up. You may want to take your board to a dealer and have them test EACH memory chip. Good-luck… – Whapping from the North Coast Michael
#12534From: Gene NeronJul 26, 1991 10:35 PM
Ok, thanks. I'll keep that in mind.